FANOVADecomposition {mlegp} | R Documentation |
Performs Functional Analysis of Variance (FANOVA) decomposition on a Gaussian process or (a subset of) all Gaussian processes in a Gaussian process list
FANOVADecomposition(gp, Interaction = TRUE, verbose = TRUE, outputs = NULL, maxpts = NULL, lower = NULL, upper = NULL)
gp |
an object of class gp or gp.list |
Interaction |
if TRUE (the default), calculates two-way factor interactions |
verbose |
if TRUE (the default), prints status updates |
outputs |
if gp is of class gp.list , a vector of integers specifying the Gaussian processes to analyze; all Gaussian processes in the list are analyzed by default |
maxpts |
the maximum number of function evaluations used in the calculation of each predicted output, see adapt |
lower |
a vector of minimum values of ALL parameters of the gp design matrix, defaulting to the minimum value of each parameter in the design |
upper |
a vector of maximum values of ALL parameters of the gp design matrix, defaulting to the maximum value of each parameter in the design |
Implements the FANOVA decomposition of Schonlau and Welch (2006), for main and two-way factor effects, using a prior distribution of all components that is ~ independent U(a,b), where (a,b) corresponds to the (min,max) value of that
component in the design. For all parameters, (a,b) can be overwritten via the arguments lower
and upper
.
The functions integrate
and adapt
from the library adapt
are used for one dimensional and multidimensional integration, respectively.
a data.frame where the first column contains the names of the effects, and elements in row i and column j correspond to the percentage of the total functional variance of Gaussian process (j-1) accounted for by effect i
This function requires the library adapt
The total contributions from all main and two-way factor interactions should never exceed 100%; however, it is possible that this will happen due to numerical rounding error
If R starts to hang, changing the default value of maxpts
is needed (though convergence no longer guaranteed)
From adapt:
Setting maxpts
to NULL asks the function to keep doubling maxpts
(starting at max(minpts,500, r(ndim))
) until the desired
precision is achieved or R runs out of memory...
Garrett M. Dancik dancikg@nsula.edu
Schonlau, M. and Welch, W. 2006. Screening the Input Variables to a Computer Model Via Analysis of Variance and Visualization, in Screening: Methods for Experimentation in Industry, Drug Discovery, and Genetics. A Dean and S. Lewis, eds. (New York: Springer).
http://users.nsula.edu/dancikg/mlegp/
FANOVAFunctions
; nice
for rounding entries in the FANOVA table; plotMainEffects
and plotInteractionEffect
for plotting main and 2-way interaction effects;
integrate
and package adapt for details on single and multi-dimensional integration
## Not run: ## Do not run the following code unless the library adapt is loaded ## ## fit the Gaussian process ## x1 = kronecker(seq(0,1,by=.25), rep(1,5)) x2 = rep(seq(0,1,by=.25),5) z = 4 * x1 - 2*x2 + x1 * x2 + rnorm(length(x1), sd = 0.001) fit = mlegp(cbind(x1,x2), z, param.names = c("x1", "x2")) ## Find percent contribution of all effects ## FANOVADecomposition(fit) ## Round contributions to 3 digits ## nice(FANOVADecomposition(fit), 3) ## End(Not run) ### end don't run